Flow analysis of centrifugal pump using CFX solver and remedies for cavitation mitigation

8/20/2019 Flow analysis of centrifugal pump using CFX solver and remedies for cavitation mitigation

http://slidepdf.com/reader/full/flow-analysis-of-centrifugal-pump-using-cfx-solver-and-remedies-for-cavitation 1/7

Dr. G. Rambabu et al. Int. Journal of Engineering Research and Applications www.ijera.com

ISSN: 2248-9622, Vol. 5, Issue 7, (Part - 3) July 2015, pp.52-58

www.ijera.com 52 | P a g e

Flow analysis of centrifugal pump using CFX solver and remediesfor cavitation mitigation

Dr. G. Rambabu*, S. Sampath** G. Karthik*** S. Siva Teja**** *(Department of Mechanical Engineering, Andhra University College of Engineering, A.U.C.E(A),

Visakhapatnam-530003

** (Department of Mechanical Engineering, Andhra University College of Engineering, A.U.C.E(A),


***Department of Mechanical Engineering, Andhra University College of Engineering, A.U.C.E(A),


*****Department of Mechanical Engineering, Andhra University College of Engineering, A.U.C.E(A),


ABSTRACT

In this scholarly thesis pertinent to the working of centrifugal pump, a CFD solver namely CFX is employed in

order to simulate fluid flow characteristics with well-defined constraints and boundary conditions defining the

problem. Stringent solid model is meticulously prepared encompassing the present day usage and constructional

features of a centrifugal pump and is constrained with various boundary conditions having fixed domain in order

to evaluate plots and results. To spearhead and facilitate this analysis program a numerical approximation tool

with high degree of convergence rate called ANSYS 15.0 software is used. The ASNYS software avoids tedious

calculations presumably impending in the design procedure and uses ultimate numerical tool to approximate thesolution of the partial differential equations associated with continuity, momentum and energy phases of a flow

problem in a 3-D model. This exquisite feature of ANSYS enables designer to optimize the design procedure inan iterative manner based on the final plots of post-processing phase. In addition, the scholarly writing also

constitutes the appraisal of the most debilitating and painstaking problem retarding the efficiency of the

centrifugal pump known as cavitation. Possible remedies for overcoming this problem will be indirectly inferredfrom the various plots and figures derived from the post -processing phase of the design process.

Keywords - Cavitation, NPSH, CFD.

I. Introduction

(What is cavitation?)Cavitation is the formation and subsequent

collapse or implosion of vapor bubbles inside the

pump. It occurs because the absolute pressure of the

liquid drops below the liquid's vapor pressure. When

vapor bubbles collapse with enough frequency, it

sounds like marbles and rocks are moving through

the pump. If larger number of vapor particles

accumulate they become irresistible and causes

indentation on the surface encompassing the pump

leading to pitting action. This pitting action corrodes

the metal leading to noisy operation of the pump and

sometimes leads to massive explosion. Cavitation is

a painstaking problem in today's commercial usage

of pump and must be avoided to avert any

retardation in the functioning of the pump.

Cavitation is bound to lower the efficiency of the pump. If the pump operates under cavitating

conditions for rough times, the following can occur:

Pitting marks on the impeller blades and on the

internal volute casing of the pump.

Premature bearing failure.

Shaft breakage and other fatigue failures in the

pump.

Pre-mature mechanical failure.

Factors perpetrating cavitation in a pump:

Fall of absolute pressure at the suction nozzle

below the saturation vapor pressure

corresponding to ambient temperature of the

water.

An increase in the temperature of the pumped

liquid.

An increase in the velocity or flow of the fluid.

Separation and reduction of the flow due to a

change in the viscosity of the liquid.

Undesirable flow conditions caused by

obstructions or sharp elbows in the suction

piping. The pump is inadequate for the system.

RESEARCH ARTICLE OPEN ACCESS






II. Estimation and prediction of cavitationA powerful tool to access the demerits of a

centrifugal pump is the net positive suction head

denoted as NPSH. This tool helps in the evaluation

of the toughest pain staking problem entangled in theoperating condition of the centrifugal pump. Net

positive suction head is defined as the minimum

head required at any point of the centrifugal pump to

avoid cavitation. It is the energy in the liquid

required to overcome the friction losses from the

suction nozzle to the eye of the impeller without

causing vaporisation. It is a characteristic of the

pump and is indicated on the pump's curve. It varies

by design, size, and the operating conditions. It is

determined by a lift test, producing a negative

pressure in inches of mercury and converted into

feet of required NPSH. According, to the standards

of hydraulic institute, a suction lift test is performedon the pump and the pressure in the suction vessel is

lowered to the point where the pump suffers a 3%

loss in the total head. This point is called the NPSHr

of the pump.

Generally, this net positive suction head theory isonly applicable to the eye and suction of the impeller

only because the vapor pressure at running

conditions is highly capricious due to the advent of

the frictional head loss and its conversion into heat

energy leading to a rise in the temperature of the

liquid. This ultimately decreases the vapor pressure.Moreover, the vapor pressure distribution is not

stagnant throughout the impeller plane and thus NPSH is not applicable to the total portion of the

pump.

+ (1)

where,

p1- absolute pressure at inlet.

pv1- vapor absolute pressure.

C- velocity at inlet.

p- density of liquid.

g- acceleration due to gravity.

and moreover,

(2)

where, NPSHr- net positive suction head required. NPSHa- net positive suction head available.

Following are the exquisite features of NPSH:

Prediction of cavitation and its evaluation.

Determining the necessary NPSH inorder toovercome cavitation at all working conditions of

the pump.

Evaluating the required NPSH by adding 3 feet

head to the available NPSH.

III. What is CFD?Computational fluid dynamics has certainly

come of age in industrial applications and academic

research. In the beginning, this popular field of

study, usually referred to by its acronym CFD, was

only known in the high-technology engineering

areas of aeronautics and astronautics, but now it is becoming a rapidly adopted methodology for solving

complex problems in modem engineering practice.

CFD, which is derived from the disciplines of fluidmechanics and heat transfer, is also finding its way

into important uncharted areas, especially in process,

chemical, civil, and environmental engineering.

Construction of new and improved system designs

and optimization carried out on existing equipment

through computational simulations are resulting in

enhanced efficiency and lower operating costs. With

the concerns about global warming and the world’s

increasing population, engineers in power-generation industries are heavily relying on CFD to

reduce development and retrofitting costs. These

computational studies are currently being performed

to address pertinent issues relating to technologiesfor clean and renewable power as well as formeeting strict regulatory challenges, such as

emissions control and substantial reduction of

environmental pollutants. The fundamental basis of

almost all CFD problems are the Navier – Stokes

equations, which define any single-phase (gas or

liquid, but not both) fluid flow. These equations can be simplified by removing terms describing viscous

actions to yield the Euler equations. Further

simplification, by removing terms describing

vorticity yields the full potential equations. Finally,

for small perturbations in subsonic and supersonic

flows (not transonic or hypersonic) these equationscan be linearized to yield the linearized potential

equations.

3.1 Computational fluid dynamics can be

summarized as follows

An approximate numerical tool used to solve 3-

D fluid flow problems whose solution hasgreater degree of convergence to the exact

solution.

An analyzer in optimizing the design procedure

by constructing its own contingent objective

function which is governed by some typicalfluid flow parameters.

A bundle of database helpful in simulating theflow characteristics of any fluid described in it.

An authentic tool which avoids the tedious,

long-winded calculations in determining the

exact solution of various partial differential

equations for a given boundary condition

stipulated by the user.

3.2 Application of CFD

In, designing Aircrafts and its aerofoiled shape

wings, tail etc for given Mach number and the

type of flow associated with it which may be

sonic, sub-sonic, super-sonic.

For, designing pumps and turbines by

simulating and imparting the desired fluid flow

conditions and characteristics.






In Estimation and optimization of a design

procedure pertinent to various fluid flowdevices and machines.

For designing and analyzing of the finned

structure encapsulating the motor cycle engine. In the appraisal and analysis of complex

structures by integrating with other simulation

analysis programs aided by the ANSYSSOFTWARE.

3.3 CFD approach and problem solving

methodology

There are generally three phases involved in the

problem solving strategy adopted by any analysis

problem. Each phase of this methodology is a

rudimentary and an unavoidable procedure for

having the desired outcome of best simulation

results predicted by the software before-hand

without an inclination to the destructive testinganalysis. The following are the three phases

associated with any ANSYS problem:

o Pre-Processing Phase:

The geometry (physical bounds) of the problem

is defined.

The volume occupied by the fluid is dividedinto discrete cells (the mesh). The mesh may be

uniform or non-uniform.

The physical modeling is defined- for example,

the equations of motion + enthalpy + radiation +

species conservation.

Boundary conditions are devised. This involvesspecifying the fluid behavior and its properties

at the boundaries of the problem. For transient

problems, the initial conditions are also defined.

o Processing Phase:

The simulation is started and the equations are

solved iteratively as a steady-state or transient.

This phase is totally dependent upon the

processing speed and type of processor

employed in the work station.

o Post-Processing Phase:

Finally a postprocessor is used for the analysis

and visualization of the resulting solution.

This phase usually constitutes various plots and

graphs contingent to the fluid flow problem being analyzed.

The solver's analysis comes under the post

processing phase of the CAD/CAM procedure after

completing the Geometric modeling and physics

definition of the problem. Actually, CFD is a

misnomer for the computing solver in analyzingvarious fluid flow problems. But the actual tool or

solver used for computing fluid flow problems is the

ANSYS SOFTWARE and its embedded tools.ANSYS is an integrated software gleamed from

various sources, database for performing wide range

analysis, optimization of design which not only

include fluid flow analysis but also convoluted problems associated with Civil structural analysis,

Heat flow analysis, Dynamic analysis, Kinematic

analysis, Electromagnetic problems etc. Varioussolvers incorporated in ANSYS for fluid analysis

are:

Fluid Flow- Blow Molding (Polyflow)

Fluid Flow- Extrusion (Polyflow)

Fluid Flow (CFX)

Fluid Flow (Polyflow)

IV. Pre-processing phase:Pre-Processing phase as described before deals

with Geometric modeling, Meshing- mesh

generation, mesh gradation, density, and lastly

comes in defining the physics of the given fluid flow

problem. The pre-processing phase serves as back bone of CAD/CAM processes and helps in providing

better visual interface to the user integrating the

graphics system along with the various periphery

devices present in the work station. A typical

workstation defining geometric modeling and mesh

generation constitutes the following items:

Graphics system and software.

Periphery devices like-plotters, mouse,keyboard, monitor, digitizer, printers etc.

CATIA software.

The pre-processing stage constitutes the crux of all

the three phases describing the 3-D geometric model

of the design process which helps in simulating the physics in the analysis stage. In addition it also helps

in gridding the complex problem to a more facile

one constituting the replica of the same element for

quick generation of results. The pre-processing stage

or phase comprises the following three indispensible

elements:

Geometric modeling

Mesh generation

Physics definition

The final facade of the drawing encompassing

present day constructional features of a centrifugal

pump is shown in the figure 4.1. The model includesan inlet, an outlet or discharge, impeller and a casing

circumventing it and is made keeping in mind the

present day constructional features of a centrifugal

pump.

Fig 4.1






V. Processing phase:Usually, the finite element approach comes

under processing phase of the design procedure

where the actual physics and boundary conditions

are devised and analysis is carried giving adequateresults. However, present day ANSYS software'ssimulating analysis program are capable of handling

all the three phases of the design procedure namely-

preprocessing, processing and post processing where

intricate 3-D models are drawn and sketched with

greater flexibility. Adding, meshing and grid

generation is also incorporated in the software.

5.1 Steps in finite-element procedures

Most generalized procedure for carrying out FEA for

any analysis problem is as follows:

Procedure 1: Discretization Divide the object of

analysis into a finite number of finite elements.

Procedure 2:Selection of the interpolation function Select the element type or the

interpolation function which approximates

displacements and strains in each finite

element.

Procedure 3: Derivation of element stiffness

matrices Determine the element stiffness

matrix which relates forces and displacements

in each element.

Procedure 4: Assembly of stiffness matrices into

the global stiffness matrix Assemble the element

stiffness matrices into the global stiffness matrix

which relates forces and displacements in the

whole elastic body to be analyzed. Procedure 5: Rearrangement of the global

stiffness matrix Substitute prescribed applied

forces (mechanical boundary conditions) anddisplacements (geometrical boundary

conditions) into the global stiffness matrix, and

rearrange the matrix by collecting unknown

variables for forces and displacements, say in

the left-hand side, and known values of the

forces and displacements in the right-hand side

in order to set up simultaneous equations.

Procedure 6: Derivation of unknown forces and

displacements Solve the simultaneous equations

set up in Procedure 5 above to solve the

unknown variables for forces and

displacements. The solutions for unknown

forces are reaction forces and those for

unknown displacements are deformations of the

elastic body of interest for given geometrical

and mechanical boundary conditions,

respectively.

Procedure 7:Computation of strains and stresses

Compute the strains and stresses from the

displacements obtained in Procedure 6

VI.

Post-processing phase:The solution, results and plots obtained in theANSYS software has great significance for the pre-

determined path chosen for analysis. Validation of

the results can be done by Rapid prototyping beforeactual advent of the designed product into the

market. In case of a civil structural problem, element

strains and stresses are the final outcomes of theanalysis along with typical plots depicting the stress

distribution along the whole continuum. The results

and plots obtained in the post-processing phase

depends upon the following factors:

Number of nodes, elements

Type of element chosen

Mesh density and mesh gradation

Number of iterations

6.1 Significance of the plots

6.1.1 Absolute pressure rotation plot (impeller with

casing)

The following adjacent plot relates the absolute

pressure distribution along the domain of thecentrifugal pump. The red portion depicts the area ofhigh pressure prone region along the volute casing of

the centrifugal pump and the navy blue portion

entails the region of low absolute pressure.

Fig 6.1.1 a

Obviously, it is true that the eye and the inlet of theimpeller i.e the suction region has low pressure

which is to be pressurized to a high pressure zone at

the discharge at outlet. The absolute pressure chart

maps the regions of the lowest pressure plots inorder

to insinuate the presence of saturated vapor pressure

at the eye of the impeller. This will serve as anindication of the cavitation in a pump. The areas of

light blue also represents the scarcity of NPSH

available at that point.

6.1.2 Pressure rotation plot (impeller pad only)

This lucidly depicts how pressure builds up along

the plane of the impeller during the running

conditions of the pump. The light blue portion

encompassing some part of the eye is more prone to

low absolute and have high risks of cavitation in

those areas.






Fig 6.1.2 a

The impeller is given a specific angular speed and

the pressure distribution plane can be observed

clearly. It must be noted that cavitation tendancy

depends greatly upon the speed of the impeller also.

During rotation the outer edge of the impeller vanes

are subjected to hydraullic stresses and muct be

backed up with intricate contours (which is nothing but the casing) inorder to sustain it.

.6.1.3 Total pressure plot (impeller back portion and

eye)

This plot gives the vulnerable areas of the high

pressure at the interspaces enclosed between thevanes. The interspaces between vanes close to the

discharge have high tendancy to hydraullic stresses

(red region) during the running condition of the

pump where as the interspaces included by the vanes

at gradually increasing cross section of the casing ismore prone to cavitation if the absolute pressure of

the liquid is below vapor pressure of the liquid at

ambient conditions. To have a better comprehensionof the total pressure plot a perspective snapshot is

Fig 6.1.3a

taken which is schematically depicted in the

folowing adjacent Fig 6.1.3 b. The hazy red portion

Fig 6.1.3 b

represents the casing encompassing the impeller and

the discharge tube at outlet. The superimposition ofthe impeller plane and casing plane gives the above

plot and can be validated as more generic form of

interpretation for judging the centrifugal pump's performance.

6.1.4 Total pressure outlet

This plot enables the manufacturers to curtail and

evaluate the performance plots of the pump

depending upon the type of application demanded by

the consumer. The difference between the inlet and

outlet absolute pressure head determines the net

Fig 6.1.4 a

effective head developed by the pump at various

speeds and operating suction head. The

superimposition of all the above plots gives the

following plot shown in figure 6.1 e which is used by the consumer in the manufacturer's catalogue for

choosing a particular type of pump for a givenapplication. The manufacturer's catalogue constitutes

a scale entailing NPSH for a given speed and

discharge head enabling the customer's to have a

greater overview of the pump's commercialexploitation.

6.1.5 Vapor superfical velocity

The most exquisite feature of CFX solver of the

ANSYS software is its ability to map and point the

areas of highest vapor velocity due to the cavitation

phenomenon causing pitting action on the casing andvane material. The vapor velocity flow pattern for a

given speed of the impeller can be easily interpreted

and comprehended. Even though the fluid attains particular pressure energy at the expense velocity

head, there will always be certain disturbances in the

flow due to the advent of cavitation leading to the

fall of pressure energy. This case is most important

pertinent to water which is making its way out of the

impeller with cavitating conditions. Consequently,

high velocity vapor bubbles start hitting the casing

encompassing the impeller, ensuing pitting action.The adjacent plot gives the vapor superfical velocity

at cavitating conditions and infers the places likely

to effect greatly by pitting action of the high velocity

vapor bubbles. The velocity of superficial vapor

bubbles greatly depends on the type of casing andthe angular velocity of the impeller. The red- dashed






lines showcase the regions with high vapor velocity

for a given speed of the impeller.

Fig 6.1.5 a

VII. Conclusion and scope: There are a number of inferences made from the

plots articulated by the CFD analysis of a centrifugal

pump. Conclusive suppositions are brought down bycontrasting the post-processing plots with different

operating conditions of the centrifugal pump. The

focus should be on resolving cavitation problems

either by increasing the external pressure on the fluid

or decreasing its vapor pressure. The external

pressure could be increased by:

Increasing the pressure at the pump suction

such that the absolute pressure of water at inlet

does not falls below the saturation pressure at

that temperature present in the ambience.

Reducing the energy losses incurring at the eyeof the pump. This can be resolved by using a

frictionless small length suction inlet. Using a larger pump.

Vapor pressure of the fluid is decreased by:

Lowering the temperature of the fluid.

Changing to a fluid with a lower vapor pressure.We are witnessing a renaissance of computer

simulation technology in many industrial

applications. This changing landscape is partly

attributable to the rapid evolution of CFD techniques

and models. For example, state-of-the-art models forsimulating complex fluid-mechanics problems, such

as jet flames, buoyant fires, and multi-phase and/or

multi-component flows, are now being progressively

applied, especially through the availability of multi-

purpose commercial CFD computer programs. The

increasing use of these programs in industries makes

clear that very demanding practical problems are

now being analyzed by CFD. With decreasing

hardware costs and rapid computing times, engineers

are increasingly relying on the reliable yet easy-to-

use CFD tools to deliver accurate results, as

described in the examples in the previous sections.

Additionally, significant advances in virtual

technology and electronic reporting are allowing

engineers to swiftly view and interrogate the CFD

predictions and to make necessary assessments and judgments on a given engineering design. Inindustry, CFD will eventually be so entrenched in

the design process that new product development

will evolve toward “zero-prototype engineering.’'Such a conceptual design approach is not a mere

flight of the imagination, but rather a reality in the

foreseeable future, especially in the automotiveindustry. Looking ahead, full-vehicle CFD models

with undershoot climate control, and external

aerodynamics will eventually be assembled into one

comprehensive model to solve and analyze vehicle

designs in hours instead of days. Time-dependent

simulations will be routinely performed to

investigate every possible design aspect. Other

related “co-simulation" areas in ascertaining the

structural integrity as well as the acoustics of thevehicle will also be computed concurrently with the

CFD models. Engineering judgment will be

consistently exercised on the spot through real-time

assessments of proposed customized designsimulations in selecting the optimum vehicle. In thearea of research, the advances in computational

resources are establishing large eddy simulation

(LES) as the preferred methodology for many tur-

bulence investigations of fundamental fluid-

dynamics problems. Since all real- world flows are

inherently unsteady, LES provides the means ofobtaining such solutions and is gradually replacing

traditional two-equation models in academic

research. The demand for LES modeling is steadily

growing. LES has made significant inroads,

especially in single-phase fluid flows. In combustion

research, LES has also gained much respectability, particularly in capturing complex flame

characteristics because of its better accommodation

of the unsteadiness of the large-scale turbulence

structure affecting the combustion

VIII. Limitations: Although much effort has been focused on

developing more robust CFD models to predict

complicated multi-phase physics involving gas-

liquid, gas-solid, liquid-solid, or gas-liquid-solid

flows, LES (large eddy simulation) remains in its

infancy of application to these flow problems.

Instead, two-equation turbulence models are stillvery prevalent in accounting for the turbulence

within such flows. LES may be adopted as the preferred turbulence model for multi-phase flows in

the future but, in the meantime, the immediate need

is to further develop more sophisticated two-

equation turbulence models to resolve these flows.

Based on current computational resources, numericalcalculations performed through LES can be long and

arduous due to the large number of grid nodal points

required for computations. However, the ever-

escalating trend of fast computing will permit such

calculations to be performed more regularly in the

foreseeable future. Also, with the model graduallymoving away from the confines of academic






research into the industry environment, it is not

entirely surprising that LES will eventually becomea common method for investigating many physical

aspects of practical industrial flows.

8.1 Following are the limitations of CFD analysis inthis project

The rate of convergence to the exact solution of

the partial differential equations is very low below 100 iterations and high speed computers

are required to enhance and interpret iterations

above 100. Generally, best solution arrives at

400-500 iterations and greatly depends upon the

type of problem chosen.

The remedies suggested for cavitation of thecentrifugal pump relents on the operational

working conditions of the pump and no

constructional optimization details is disclosed

by the project thesis. Cavitation analysis made in this project is

confined to both on-load and off-load phases.

However, it is difficult to infer conclusions

from the on-load characteristics because it is

contingent upon various factors like type of the

casing material, vane material, thermal

properties of the vanes and casing adhering tofrictional head loss resulting in highly

capricious NPSHr (net positive suction head

required).

No methods or remedies are studied regarding

the erroneous vibrations and noise emanated

from the working of a centrifugal pump.

IX. Nomenclature:CFD- computational fluid dynamics.

NPSH- net positive suction head.

NPSHa- net postive suction head available.

NPSHr- net positive suction head rquired.

p1- absolute pressure at inlet.

pv1- vapor absolute pressure.

C- velocity at inlet.

p- density of liquid.

g- acceleration due to gravity.

X.

Acknowledgment:This project thesis and documentation is done

with grim determination under strict tutelage of our

mentor Dr. G. Rambabu and any claims raised by

the third party for a facsimile unauthenticated copy

is merely a coincidence and will be disavowed by

the members of the batch. In addition, any attempts

for emulating this thesis content and analysis will

also be condemned and pertinent legal action will be

sued against the third party. Finally we acknowledge

the contribution of all those who have helped us

directly or indirectly with their good wishes and

constructive criticism which lead to successfulcompletion of our Project.

References:

Journals: [1] Reddy, J.N. (2005). An Introduction to the

Finite Element Method (Third ed.). McGraw-

Hill. journal.[2] Hrennikoff, Alexander (1941). "Solution of

problems of elasticity by the framework

method". Journal of applied mechanics 8.4:

169 – 175.[3] Elman; Howle, V; Shadid, J; Shuttleworth, R;

Tuminaro, R et al. (January 2008). "A

taxonomy and comparison of parallel block

multi-level preconditioners for the

incompressible Navier – Stokes

equations". Journal of Computational Physics 227 (3): 1790 – 1808.

[4] Courant, R. (1943). "Variational methods for

the solution of problems of equilibrium andvibrations". Bulletin of the American

Mathematical Society 49: 1 – 23.

[5] Hirt, C.W.; Nichols, B.D. (1981). "Volume of

fluid (VOF) method for the dynamics of free

boundaries". Journal of Computational

Physics.

Books:[6] CAD/CAM applications by P.N.Rao. [7] Shepard, Dennis G. (1956). Principles of

Turbomachinery. McMillan.[8] Reti, Ladislao; Di Giorgio Martini, Francesco

(Summer 1963). "Francesco di Giorgio(Armani) Martini's Treatise on Engineering

and Its Plagiarists".

[8] Fluid mechanics and hydraulic by Modi and

Seth [9] Centrifugal pump design by John Tuzson

[10] Hirt, C.W.; Nichols, B.D. (1981). "Volume of

fluid (VOF) method for the dynamics of free

boundaries". Journal of Computational

Physics.

[11] Unverdi, S. O; Tryggvason, G.(1992). "AFront-Tracking Method for Viscous,

Incompressible, Multi-Fluid Flows". J.

Comput. Phys.[12] Benzi, Golub, Liesen (2005). "Numerical

solution of saddle-point problems". Acta

Numerica 14:1 – 137.

.

http://en.m.wikipedia.org/wiki/Gretar_Tryggvason




Documents

Flow analysis of centrifugal pump using CFX solver and remedies for cavitation mitigation